Low-dose CT screening may lead to lung cancer overdiagnosis in low-risk patients
Click Here to Manage Email Alerts
Key takeaways:
- Low-dose CT imaging may correlate with lung cancer overdiagnosis in China.
- Increases in early-stage incidence observed with small decreases in late-stage incidence suggests overdiagnosis.
High incidence of lung cancer paired with stable lung cancer-related mortality rates when low-dose CT screening was introduced suggests overdiagnosis, specifically in women, according to study results published in CHEST.
2011 marked the start of China’s implementation of low-dose CT (LDCT) screening in employee health examinations, according to researchers.
“Our findings of a rapid increase in incidence with little change in mortality and a significant increase in early-stage incidence not accompanied by a substantial decline in late-stage incidence since 2011 in Chinese women provide evidence for LDCT imaging-induced lung cancer overdiagnosis in the population,” Mengyan Wang, PhD, of the department of epidemiology at Fudan University Shanghai Cancer Center, and colleagues wrote.
In a population-based study, Wang and colleagues analyzed 34,152 patients with incident lung cancer from 2002 to 2017 from the Cancer Surveillance of Shanghai to determine if LDCT screening has contributed to lung cancer overdiagnosis in low-risk populations.
Researchers figured out age adjusted rates of incidence, both stage-specific and histologic type-specific, and mortality in men and women to assess temporal trends of both factors and determine if they signaled overdiagnoses.
Of the total cohort, 27,208 deaths due to lung cancer were observed.
When adjusted for age, researchers found that men had a lower incidence of lung cancer from 2002 to 2008 (annual percentage change, –2.7%; 95% CI, –4.91% to –0.43%) and death related to lung cancer from 2002 to 2009 (annual percentage change, –2.93%; 95% CI, –4.6% to –1.22%). From both 2008 and 2009 onward, there have only been small annual percentage changes (incidence, 1.07%; 95% CI, –0.18% to 2.34%; mortality, –0.67%; 95% CI, –0.27% to 0.75%).
However, for women, 2011 marked the start of a rise in lung cancer incidence (annual percentage change, 11.98%; 95% CI, 9.57%-14.45%). Notably, mortality in women continually went down during the study period (annual percentage change, –1.27%; 95% CI, –2.12% to –0.41%). Researchers wrote that when both factors do not go in the same direction or one remains stable, overdiagnosis may be present.
Stage-specific incidence
When assessing stage-specific rates of incidence over 2002 to 2017, men had a smaller increase in incidence of early-stage cancer cases than women (absolute difference, 6.9 per 100,000; 95% CI, 5.1-8.7 per 100,000 vs. 16.1 per 100,000; 95% CI, 14-18.3 per 100,000), as well as a larger decrease in incidence of late-stage cancer (–5.5 per 100,000; 95% CI, –9.2 to –1.7 per 100,000) compared with women, who showed no significant decrease (absolute difference, –0.6 per 100,000; 95% CI, –2.8 to 1.7 per 100,000).
For women, this combination of high early-stage incidence and no changes in late-stage incidence indicates overdiagnosis, according to researchers. Importantly, researchers wrote that lung cancer overdiagnosis might be present in men as well based on these stage-specific findings.
Additionally, both men and women saw lower mortality over the study period (absolute difference men, –11.3 deaths per 100,000; 95% CI, –16.4 to –6.2 deaths per 100,000; women, –3.5 deaths per 100,000; 95% CI, –6.4 to –0.6 deaths per 100,000) and higher 5-year survival rates from 2002 to 2014 (men, 9.9% to 18.5%; women, 8.5% to 38.5%).
Histologic type-specific incidence
For both men and women, lung adenocarcinoma showed the greatest increase in incidence among the histologic types. Women went from 5.9 cases per 100,000 in 2002 to 30.4 cases per 100,000 in 2017, and men went from 7.7 cases per 100,000 in 2002 to 21.6 cases per 100,000 in 2017.
Researchers also found cancer of unknown histologic type went down from 2002 to 2017 in women (difference, –4.9 cases per 100,000; 95% CI, –7.3 to –2.5 cases per 100,000) and in men (difference, –13.9 cases per 100,000; 95% CI, –17.8 to –10 cases per 100,000).
Incidence rate ratios
Lastly, evaluation of incidence rate ratios by sex demonstrated a significant increase from 2002 to 2017 in women (average annual change [AAC], 0.05; 95% CI, 0.04-0.06), with the greatest increases found in women aged younger than 45 years (AAC, 0.11; 95% CI, 0.03-0.19). Additionally, stage I lung cancer (AAC, 0.21; 95% CI, 0.17-0.25) and adenocarcinoma (AAC, 0.12; 95% CI, 0.1-0.14) had the highest AACs when separated by stage and histologic type in women. In terms of significant decreases in these categories, researchers found them in stage III lung cancer (AAC, –0.04; 95% CI, –0.05 to –0.03) and squamous cell carcinoma (AAC, –0.05; 95% CI, –0.09 to –0.02).
According to researchers, these results again show the increase in early-stage cancer and lung adenocarcinoma in women from the incidence-specific and histologic type-specific analyses.
From 2002 to 2017, the overall incidence rate ratio did not change in men, but they did show similar patterns of rising in stage I cancer (AAC, 0.08; 95% CI, 0.05-0.11) and adenocarcinoma (AAC, 0.08; 95% CI, 0.06-0.09), as well as falls in squamous cell carcinoma (AAC, –0.03; 95% CI, –0.06 to –0.01).
Although Wang and colleagues show evidence of possible overdiagnosis by LDCT in China through a combination of high lung cancer incidences with little mortality change, they acknowledge that tobacco use, air pollution and a low-fruit diet could also be behind the increase in lung cancer incidence.
“Criteria for LDCT screening in low-risk populations and management of screening-detected nodules need to be addressed fully for the expanded application of LDCT imaging in lung cancer screening,” Wang and colleagues wrote.
This study by Wang and colleagues demonstrates potential dangers that come with screening low-risk individuals for lung cancer, which in turn supports the U.S. screening recommendations, according to an accompanying editorial by H. Gilbert Welch, MD, MPH, of the department of surgery at Brigham and Women’s Hospital, and Gerard A. Silvestri, MD, professor of medicine and lung cancer pulmonologist at the Medical University of South Carolina.
“The current United States Preventive Services Task Force recommendations only support screening those 50 to 80 years of age with a 20 pack-year smoking history,” Welch and Silvestri wrote. “They do not support screening individuals who never smoked. Modeled lung cancer risk estimates for 65,711 individuals who never smoked consistently was less than the threshold where screening would be beneficial. Thus LDCT scan screening should continue to be restricted to those at highest risk, while rigorously maintaining adherence to growth assessment protocols.”